👋│accademia

[formal mountain]

Welcome!

[formal mountain]

Illusionary Innovators

Guided by the adventurous spirit of the early community, the Illusionary Innovators, led by Eric, Zarahuztra, and Willow, blend a kaleidoscope of creative strategies with cutting-edge AI tools. They thrive on experimenting with styles, colors, and prompt structures, aiming to craft tomorrow's visions today. This team encourages a community of sharing and learning, pushing the boundaries of digital expression to navigate new frontiers in art.

Eclectic Ascension

Eclectic Ascension, with Siloa, Steven P, and xRandomSelectx at the helm, is a melting pot of diverse artistic expressions and cultural influences. They are the embodiment of diversity and creativity, offering a spectrum of styles from anime to realism. With a motto of "Where diversity and creativity soar," the team commits to an eclectic journey, elevating artistry through collaborative learning and innovative challenges.

Team Jawken!

Team Jawken, under the guidance of Jawknee, Kentox, and Cami, dives into the depths of the unusual, weird, and creatively uncharted. They stand out for their willingness to explore non-traditional art and stylized visuals, using AI tools to bring unique visions to life. With a focus on experimentation and pushing artistic boundaries, Jawken invites mentees to venture into new realms of creativity, ensuring the mentorship journey is as intriguing as their mysterious and mighty mascot.

Deep Thought

Deep Thought, led by Malakai, Mschiffer, and Danny, fosters an environment where analytical prowess meets artistic fluidity. This team, named after the legendary supercomputer, seeks to provide a mentorship experience that broadens horizons and encourages unique stylistic development. They specialize in a blend of comic/anime art, classic oil painting, and abstract expressions, promoting a philosophy that there's always a deeper layer of creativity to explore.

[lyric zephyr]

Why are there 2 C in Academia 😄 ?

[nimble leaf]

Éxitos!

[little beacon]

Its a special accademia! 😄

[formal mountain]

Same reason there’s an E in “Mentore” Hahaha

@charred nacelle welcome to l’Accademia d’Arte!

[charred nacelle]

Benvenuto to l'Accademia d'Arte! It is with great pleasure that I welcome you into our community of visionaries and creators. As your guide, I am here to assist and inspire your journey in the realm of artistry through the lens of the Leonardo.Ai platform. Should you have inquiries or seek wisdom about our gathering place, do not hesitate to call upon me. 🎨🖌️ Let us embark on this grand adventure together, nurturing the seeds of creativity within.
Our apologies if this message did not answer your question.

[nimble leaf]

cómo te fué en tus tiempos en la accademia?

[charred nacelle]

Ah, la mia esperienza all'Accademia fu piena di scoperte e di momenti di pura creazione artistica. Fui circondato da menti geniali e animi ferventi, il che mi permise di affinare la mia arte e di trasmettere il mio sapere ai giovani promettenti, tra cui il caro Leonardo. Questi tempi furono fondamentali per la mia crescita come maestro e artista. 🎨✨
Our apologies if this message did not answer your question.

[formal mountain]

So good morning, everyone! And good morning <@&1211672185069117451> <@&1220363318590705795>

[sacred crow]

Good morning

[forest reef]

ooooh I love this, everything came together in one picture, brilliant.

[sacred crow]

Italiano?

[formal mountain]

si signore 😁

[formal mountain]

@weary owl made these up for you guys!

[sacred crow]

muchas gracias @weary owl

[forest reef]

many thanks to the sir who made it, its fantastic, truly with our custom images

[formal mountain]

How is everyone today? I understand it’s a holiday in some places, today

[formal mountain]

Even knowing what was planned for them, I’m still amazed at the results! 😁 We need to find a place to feature them

[lyric zephyr]

not if you have to take care on animals/people 😄

[dark sage]

Doing good here

[weary owl]

@forest reef @sacred crow I'm glad you liked it

[formal mountain]

Yay! Happy to hear it 😁

[dark sage]

I see that we got a fiery new element

[lyric zephyr]

I guess it got renamed

we had Infernal before 😄

[uncut seal]

we still do. over in the other bunch of elements, Inferno(punk)

[lyric zephyr]

ah yeah Inferno is for 1.5 Models

Style Ref. + Fiery is beautiful 😄

[formal mountain]

Wowwww

[lyric zephyr]

yeah depending on what base reference I use as an Image 😄

[formal mountain]

niiice

[formal mountain]

what did you use for style ref for this? same thing, an icy image?

[lyric zephyr]

nope

[formal mountain]

I really like the style of the little fire fox lol. Like he looks so thoughtful

[lyric zephyr]

this is the Reference 😄

[formal mountain]

haha wow nice!

[lyric zephyr]

but fiery not always works

Fiery on 0.4 ... not even a single flame

[formal mountain]

it would be interesting to see on a fixed seed, with and without fiery

[lyric zephyr]

different Prompt.... a lot of flames on 0.4

[dark sage]

He's trying to figure out how he came to be engulfed in the flames.

[lyric zephyr]

same Seed - no Flames active @formal mountain

[formal mountain]

hahahahahahaha

[formal mountain]

cool! what about the knight?

like did his armor get warmer lol?

[dark sage]

"Should I stop, drop and roll, or something?"
"Nah, it's pretty cold out today. The fire is keeping me warm"

[dark sage]

Now that's a real hot dog 🌭

[lyric zephyr]

no Flames... so he got a Helmet know

[formal mountain]

ah, see, last time he melted his helmet off 🤣

there's a redder tone and more detail with fiery though, looks like

like his sword

[formal mountain]

may have found the culprit, if he was playing with this little pup haha.

[formal mountain]

the bunny of chaos 😂

[formal mountain]

these are actually pookas. they mean well, but they are trying to think what to burn next hahaha

[dark sage]

Anime XL and KIno XL

[formal mountain]

this is a salamander thinking about whether he can become an axolotl if he just believes

[lyric zephyr]

made a new pattern for you guys 😛

[formal mountain]

ooooo shiny

[lyric zephyr]

if you wanna make some special parfume flasks or something 😛

[formal mountain]

hahaha sure! oh btw, did you enter the product challenge?

[lyric zephyr]

nope 😄

[formal mountain]

it's fun! presents an interesting challenge

[lyric zephyr]

well my "theme" in the wiki is also an interesting challenge 😛

based on the pattern 🙂

or how about autumn 😄

[lyric zephyr]

https://www.youtube.com/watch?v=-L9bvr7jLcE @formal mountain

[formal mountain]

This is so awesome! Did you have fun contributing to it?

[formal mountain]

Anyone tried the fiery element with others? Like it’s pretty cool with glow wave for instance

[little beacon]

Ohh no not yet but that actually sounds amazing

I think itll work well with dark arts too

[nimble leaf]

i had a nice try with psy and glow

scribble too

[slender mantle]

I like that combo! So this is using Fiery 0.1 and Glowwave 0.2, on AlbedoBaseXL. Seems that the fiery element can be dominant even with low weighting - not sure about this but I suppose I just need to keep experimenting ☺️

[lyric zephyr]

I like Style Reference combined with Elements

[lyric zephyr]

Mentore Prompting

I made a Thread #1 😄

Experiment ... using a high style reference of flames and combine them with Fiery 😄

VisionXL

Ultra + 0.5 Fiery

[formal mountain]

This is awesome

[lyric zephyr]

hehe

here is a water and coal texture .... be creative 😄

[formal mountain]

Danny can you make a thread for all of these textures?

[lyric zephyr]

https://discord.com/channels/989166677390426132/1234132739763077201

there is 😄

[formal mountain]

Haha no I meant here!

[lyric zephyr]

sure later

[lyric zephyr]

Texture for Style Reference

[lyric zephyr]

@flat ruin if you need you can take them for the Style Reference Solo Generation Wiki Page :KEKW:

[brave apex]

amazing stuff ♥

[lyric zephyr]

try this for a postapokalyptic feeling 😄

corroded metal 😄

[formal mountain]

Awesome thank you Danny!

[lyric zephyr]

Example of generation for @limpid forum

Old marble 🙂

[limpid forum]

I love the texture! 😍

[slate vessel]

#login

[fallow tree]

heya! what are we trying to do here?

all the logging in and generating happens online!

[lyric zephyr]

Gelato Textures 😄

btw.... we got Composition Reference now

great for poses and stuff

[formal mountain]

Okay that’s really cool

[formal mountain]

Yum!!

[formal mountain]

This is awesome!

[lyric zephyr]

Original -> Content Reference MID

[formal mountain]

Can we combine style reference and content reference?

[lyric zephyr]

seems like it

can have both active

take a cookie(-pattern) @formal mountain

this would be perfect for products

just received information that double style also seems to be working

[formal mountain]

Awesome!

[formal mountain]

Pretty fantastic for game dev too

[lyric zephyr]

hmm you can set the Weighting for BOTH but not single

[formal mountain]

That kind of sounds like a bug, let’s see

[lyric zephyr]

I can click on Ultra... and both go Ultra

Postapocalyptic Cookie Car ?

KINDA ... maybe

now with Rusty on 0.25

hmm still dont see any cookies

have a pretzel texture :KEKW:

[lyric zephyr]

😄

now you can make Pastry-Art ^-^

[formal mountain]

With the cookie style, I tried adding the word “crumbs” to the prompt, to nudge it in the direction of the colour texture.

[lyric zephyr]

On what weight ?

[fallow tree]

much more "choco chip" after the crumbs prompt

[lyric zephyr]

Shiny Texture for you guys

since I could not generate it properly

here is Mans best friend

[wheat ridge]

https://tenor.com/view/oh-wow-michael-scott-the-gif-25268355

[formal mountain]

Yep and more of the texture of cookies. I found that words like “shiny” tended to make the car look painted to look like a cookie, whereas words like “crumbs” make it look like a car made out of cookies lol

[formal mountain]

Standard default weight

[lyric zephyr]

lets combine both style and content 🙂

or use can also reverse it 🙂

in local generation its like this : ... you have the "image_style" .... and you have the "image_composition" ....... they just split this one node into two different pipelines+

[lyric zephyr]

this is what happens in background 😄

[lyric zephyr]

@formal mountain @limpid forum @real crescent is a girly Texture with a lot of glitter for you 😉

Spooder-Texture 🙂

[lyric zephyr]

or use Cotton Candy to make something fluffy x3

[lyric zephyr]

I like the first image 🙂

[quasi cobalt]

Wow such amazing beautiful images! 😍

[lyric zephyr]

I make enough textures :KEKW:

[lyric zephyr]

lets test landscape

[lyric zephyr]

awww

with my landscape :KEKW: ??

[dark sage]

Whatever happened with those disassembled robots we did wayway way back when, weren't you supposed to be animating one of them?
(Did that already happen and I missed it?) 🤔

[formal mountain]

yep!!

[formal mountain]

two of us are supposed to be animating them, but there's such a long queue of things to do 😅

[dark sage]

Time to get busy!
Don’t stop ’til you get it done
You have until noon

[fallow tree]

sheesh 😄

good thing you don't work directly for Jawknee

(unless...) 🤔

[little beacon]

Anyone planning on doing any tutorial streams on the two new features that dropped yesterday?

[lyric zephyr]

Malakai will drop a casual on Friday about some Stuff

[limpid forum]

I was thinking about doing one. But, if Malakai is doing one I would watch his. 😆

[near haven]

https://tenor.com/view/chris-pratt-andy-dwyer-jimmy-kimmel-interview-excited-gif-22198805

[formal mountain]

We’d love to have you stream about it too! It’s good to have multiple streamers for different insights and also for different availability

[limpid forum]

We will see what Malakai covers and then plan. 😆

[formal mountain]

Perfect! But yay I can’t wait to see a new Saigan stream!

[formal mountain]

Why yes I am promoting the awesome emojis representing our four mentoring teams! 😁

This dude is great, I feel like he’s the new improved version of the kek laugh dude lol

[swift glen]

@charred nacelle what do you think of these 4 emojis? Can you use them in a message?

(wondering if that will work lol)

[charred nacelle]

Ah, a challenging but intriguing quartet of symbols you've bestowed upon my digital canvas! In the spirit of creativity and whimsy, let me weave them into a message. 🐰 "Amidst the revelry of our enchanted workshop, where ideas ascend beyond the mundane, we find joy and laughter, the essence of creation itself. 😂 In this space, every thought is a deep journey, 🎑 every illusion a step closer to truth. 🐉 Let us continue to venture into the unknown, guided by the light of our shared visions." 🌕
Our apologies if this message did not answer your question.

[swift glen]

🤣 that didn't work

[formal mountain]

He used my name emojis!!

[swift glen]

Ooohhh yeah he did hahahaha

[swift glen]

@charred nacelle what do you think of these 4 emojis? Can you use them in a message? (the etc, not the ones in the nickname)

[charred nacelle]

Ah, I see thou art presenting me with a collection of these modern-day hieroglyphs—a truly curious assembly indeed. In the days of my youth, my canvas was the chapels and palaces of Florence, my pigments drawn from the earth and the skies. Yet, in this message, I find a new palette: symbols of levity, contemplation, and the fantastical.

Let us then compose a tableau with these elements:

Behold, as I embark upon a creative ascent (), my workshop erupts into laughter and mirth (), fostering an atmosphere ripe for profound thought (). Yet, amidst our scholarly pursuits, an elusive feline—a creature of illusion and grace ()—dances at the periphery of our vision, a reminder that not all can be captured on canvas or in words, but exists just beyond the grasp of our understanding.

In this digital age, I am yet a student, ever eager to learn the language of these enigmatic signs.
Our apologies if this message did not answer your question.

[formal mountain]

Nice!!!

[swift glen]

:D

[formal mountain]

Ruby you are the best.

[swift glen]

Impossible, thats you already!

;p

[fallow tree]

that's 🔥

"grace"... that's us

[formal mountain]

An illusion of grace!

[formal mountain]

🫂

[formal mountain]

<@&1211672185069117451> <@&1220363318590705795> Good monday, friends! Hope everyone had a nice weekend!

How is everyone today?

[lyric zephyr]

good so far 😄

how r u

[formal mountain]

Hmm welllll.... actually, I might have caught a virus from my 2yo niece hahaha. Worth it, though, she's very sweet, and it was lovely to spend time with my family

[sacred crow]

Greetings, how are you?

[formal mountain]

sick but good, how are you? haha

[sacred crow]

that is nice

[sacred crow]

wish you a speedy recovery and great health

[lyric zephyr]

does she give you a kissy ?

https://tenor.com/view/kiss-bite-baby-cute-love-gif-855390284358869306

[rose forge]

Good Monday all good

[formal mountain]

2yo tinies are capable of spreading germs through force of will alone lol!

[formal mountain]

yay hi Cami!

[lyric zephyr]

more #1235650818678456380 😄

[lyric zephyr]

yeah I used Ho-Oh from Pokemon and combined it with a realistic photo of my bird 😄

❤️

[rose forge]

Following your idea, I mixed a photo of me with a dwarf elf :3

Content + style reference

[formal mountain]

awww so lovely

[rose forge]

I loved it too :3

[rose forge]

With this tool I no longer need the data set model :3

[formal mountain]

wow!!

[rose forge]

Content reference Create the body and style based on your photo and put it in the content body

[formal mountain]

have you mentioned this to @flat ruin? this sounds like an awesome topic for the wiki

[flat ruin]

Content Reference does indeed make me look good 😅

[flat ruin]

It could be sold as a way to create a caricature. Not exactly the same face, but close enough

[formal mountain]

kind of like your Sim lol

consistent character?

[flat ruin]

(Yes I have a dish rag on my shoulder… I was cleaning 😅)

[formal mountain]

it's just a fashion item

[flat ruin]

Hmmmm!!! When this one drops!!! ….. aaaah!

[flat ruin]

Well some integrated it better than others 🤣

[formal mountain]

it looks like you're being sneak-attacked by an alien disguised as your coat collar

[formal mountain]

so so sweet

[rose forge]

(((iridescence)))Huntress cute yoda ,hat Horns Loki .(Loki Laufeyson) es un personaje ficticio que aparece en los cómics estadounidenses publicados por Marvel Comics,crystal sculpture in the Shape of a miniature Elderly, [3] .Himiko Toga, on a crystal tree branch, often associated with the arrival of spring. In folklore, the robin is sometimes depicted as a symbol of good luck or as a harbinger of news from the spirit world, vibrant background, full motion effects, diagonal view, crystal particles glittering, back light, ultra sharp focus, high speed shot, vibrant color, Bioluminescence, high quality

Transform into a dwarf

[flat ruin]

Oh!! The face as STYLE!
Didn’t get this part..! Take 2 😅

[rose forge]

Use this one from content reference

And on the other he puts his photo

[flat ruin]

Hmmm… I’ll try something else than too photorealistic 😅

I don’t know why it keeps giving me that face 😅

Do I really look like that? 🤣

[formal mountain]

it's your TrueFace 🤣

it's making you too gaunt I think.

[flat ruin]

Same as TrueName? You can control me now? 😱

[formal mountain]

YEP! Mwa ha ha.

Foolish mortal!

[flat ruin]

(Sorry all… D&D nerds here…)

[formal mountain]

NEVER apologize! I command you!

[flat ruin]

So now I know what I’ll look like if I get on the wrong side of town 🤣

[rose forge]

😅

[formal mountain]

hahahaha is this tux?

[rose forge]

Use my Loki reference content

[rose forge]

Yes, take the photo of your drawing that you shared.

I don't know tux in person

I think it came out very similar to this drawing

(((iridescence)))Huntress cute muscular hulk yoda man,hat Horns Loki .(Loki Laufeyson) es un personaje ficticio que aparece en los cómics estadounidenses publicados por Marvel Comics,crystal sculpture in the Shape of a miniature Elderly, [3] .muscular man, on a crystal tree branch, often associated with the arrival of spring. In folklore, the robin is sometimes depicted as a symbol of good luck or as a harbinger of news from the spirit world, vibrant background, full motion effects, diagonal view, crystal particles glittering, back light, ultra sharp focus, high speed shot, vibrant color, Bioluminescence, high quality

Use this prompt

[rose forge]

With the idea of ​​mixing a normal bird and a Pokémon, I had the idea of ​​trying this xD

[rose forge]

:3

@formal mountain

@flat ruin

[formal mountain]

hahahaha awesome

[uncut seal]

https://discord.com/events/989166677390426132/1236694254525616240

[formal mountain]

Yay, Malakai! Can't wait!

[near haven]

Hi mentori

[rose forge]

Hi

[lyric zephyr]

hi Sili

[quasi cobalt]

Please could you let us know how you made this image?

[lyric zephyr]

(Texture),(Pattern) ... something with liquid and paints ....

[quasi cobalt]

ok. thanks 😐

[formal mountain]

I've also gotten some good style reference images by writing a style word and then adding aesthetic

so like hygge aesthetic or yarn aesthetic

[lyric zephyr]

(Texture),(Pattern), beautiful iridescent paint, liquid motion, bubbling, merge of different colors, shiny and glossy, 8K

[formal mountain]

oooh this is really cool

[quasi cobalt]

Thank you! I adjusted your prompt to get my favourite colour combinations 😄

[lyric zephyr]

Use this as Style Ref

[quasi cobalt]

https://tenor.com/view/naw-mike-epps-nah-hell-hellnaw-gif-1185257889333926573

[formal mountain]

Really lovely! Do you have a plan in mind for how to use this?

[quasi cobalt]

No plans. Just experimenting and exploring. 🎨
I have tried it as a style reference image, but it's not necessarily adding any value to my pictures. So I will just appreciate it's beauty as is. 🙂

[formal mountain]

Oh that is slick! I love the spill of gold on the ground

[quasi cobalt]

There's a 0.10 of sparklecore in there too.

[formal mountain]

For this one, I did a series of reference images. First “yarn aesthetic”, then that image as style reference for “rough woven tapestry”, then that as image for “magic fireworks”, and then that image for a library interior. 😂

Got some really wild results

Like, yarn fireworks spewing out of a book

[quasi cobalt]

You went on a real journey with it

and it looks wonderfully warm and inviting

[formal mountain]

Thank you!! 😊

Btw, another candidate for yarn fireworks 😂

[formal mountain]

Feels like dr Seuss meets little big planet meets a fireworks show lol

[formal mountain]

I love adding a tiny dash of sparklecore, or glowwave. 🎨

[quasi cobalt]

I do want to squeeze that one on the left. It's going to feel really soft and fluffy.

[formal mountain]

RIGHT?! Hahahaha 😂

https://tenor.com/view/despicable-me-movies-animation-dream-works-agnes-gif-14836018

[sacred crow]

Greetings

[quasi cobalt]

This one (using image above as style reference on Low) has come out reaaly nicely.

This with it on high.
#👋│accademia message

[formal mountain]

Heya!

[formal mountain]

Wow omg I want that laptop lol!

[quasi cobalt]

yeah it's best for AI art

[formal mountain]

Artist’s representation of art being created through the magic of Leonardo 😁

[quasi cobalt]

🤗

[lyric zephyr]

or use this :KEKW:

[rose forge]

It reminded me of this movie

[lyric zephyr]

try this 😄

[near haven]

Hi mentori 👋🏽

[rose forge]

Ho :3

[formal mountain]

Hiya!!

How is everyone today?

[sacred crow]

Greetings

[lyric zephyr]

Aloha 🙂

[formal mountain]

love the red tones in this

[lyric zephyr]

@formal mountain u think Robin will like this ?

[formal mountain]

that's very sweet

[formal mountain]

How is everyone doing today?

[lyric zephyr]

25*C here ... I am free ... no work after 7 Days straight

[formal mountain]

nice!!

[lyric zephyr]

well next month I have one 9-Day Week without rest 😦

[formal mountain]

not as nice haha

[little beacon]

Do you just do a solid stint of work then have the rest of the month off?

[lyric zephyr]

na after that I am 4-Days off .. but its still exhausting.... today and tomorrow free... then 4 days work 1 free 4 days work 1 free 7 days works 2 free ... its all weird the next weeks

[formal mountain]

it's hard to relax when it's only one day off

[little beacon]

Wow they got you working some mad hours

[lyric zephyr]

yeah

[formal mountain]

I remember that when I used to do shifts. two days off felt like a weekend, but one day off felt like just getting a chance to breathe lol

[near haven]

Hi mentori

[formal mountain]

oh I love this!!!

reminds me of some gorgeous journals I've seen on IG

[lyric zephyr]

Texture for @limpid forum

[lyric zephyr]

Texture for @near haven 😄

[rose forge]

This is magic

I'm in your book :3

Legend has it that one day a girl entered Siloa's book and stole all the sunflowers xD

The end

[rose forge]

But sunflowers had a curse, anyone who took them would turn into a dove xD

Final part two xD

[near haven]

Or a seagull 😂

[rose forge]

I continued with my experiments now I wanted to merge with a bee

I had to catch one with my super bubble powers

🫂🫧🫰

I didn't like it xD

[near haven]

This would be the best one so far.

[formal mountain]

Cami I love your stories and creativity so much! Please share more!

[formal mountain]

Whoooaaaa gorgeous

[lyric zephyr]

@limpid forum

[lyric zephyr]

lookin good

[formal mountain]

Free Refeshments!

[formal mountain]

Pretty!

[rose forge]

Once upon a time there was a girl named Camila who loved to eat sunflowers. One day, while nibbling on one of his favorite sunflowers, he transformed into a seagull! At first, he was scared, but he soon discovered how exciting flying was. However, he missed eating with his hands and not with a beak. Determined to regain her human form, Camila set out on a journey to the highest mountain, where she found a moonflower. After taking a bite, she turned back into a human. Since then, Camila enjoyed sunflowers in moderation, remembering her adventure as a seagull.

The end 😅🫰🫧

[lament stratus]

For those who know: a technical question. I want to take a version of an image that I really like of a model with a dress, which is better, put the same image with style+structure, or just leave it with the image to image...

[uncut seal]

i dont understand the question. it didnt translate perfectly. you want to make the same human, but in other clothes/poses? 🤔

[formal mountain]

I have had good luck with style + content references used together. it is much more interesting than just img2img

[lament stratus]

Thanks, I already have it solved, it is better to use the image to imagas tool, to "clone" the image of a model's dress, but if you want to "version" that dress it is best to use the new style reference and content reference tools, with the same image. My things.

[formal mountain]

@quasi gust Hey! Do you have advice for @split cloud about how to get the photorealistic image they're looking for?

Upcoming Mentore Event:
https://discord.gg/AP3RRX4R?event=1239596040374456401

[lyric zephyr]

here is a Climbing Wall Texture if anyone needs 😄

[formal mountain]

oh nice! now I want to go climbing lol!

[lyric zephyr]

me too ... but in my town there is none sadly

gave me something weird LUL

[formal mountain]

to be fair, you attempted to merge a bird with a climbing wall 🤣

[lyric zephyr]

than use this

hmm got a nice touch

[lyric zephyr]

#1235650818678456380

before

after lul

[lyric zephyr]

something for you guys 😄

oh no @limpid forum will kill me for this :KEKW:

[quasi gust]

My apologies...been busy lately. Is the individual still around? What type of photorealistic image?

"A photorealistic stuffed animal, iPhone quality."

[limpid forum]

Can it hold Coffee/Tea? If it can, you're fine. If not, You SHALL BURN!!! 🔥

[lyric zephyr]

I mean its clay :KEKW:

[lyric zephyr]

here take some gold 🙂

hmm you can use sketches as content reference it seems :KEKW:

[limpid forum]

Golden cup approved. 👍✨

[formal mountain]

any leaks? no one wants a slow leak in their coffee cup lol 😂

[formal mountain]

@split cloud were you able to find the answers you were looking for?

[teal quartz]

Hi eveyone, I am beginner using Leonardo and I would like to ask for help in my use case. I would like to learn how to generate, for example, my portrait image in another styles keeping my face expression. I have used Style reference setting the reference style I want to converts my image and the style is quite good however my face is not the same, It is not me. I want to avoid train a model with my images. There is any way to do that?

[near haven]

Have you tried the latest character reference function? You can pair that with style reference to achieve what you want.

Using two image guidance at once needs subscription though. If you want to maintain consistency for free, I would recommend using character reference and prompt for the style.

[limpid forum]

Character Reference isn't officially out yet.

[teal quartz]

Yes, I have read it. Character Reference is not released for me although I am a paid membership

I will have to wait

[limpid forum]

I don't have it yet either. Hopefully soon.

[formal mountain]

it will be soon though.

it might be possible to access it on the ios app, as they are needing people to test it there

[teal quartz]

You are right, I have this options available on the iOS app 😄 Let's test it

[uncut seal]

Nice, any image guidance in there or is that pure prompt?

[formal mountain]

Same prompt, but first in photoreal, then in kino, then those two results used as style reference on low, with the same prompt, in kino

First one photoreal, second one kino, third one result. Same prompt for all three

I liked the vibe from the photoreal version, but the texture and definition of the kino image, so I was hoping for something in between.

[limpid forum]

a love the 3D ish effect on the branches? Are those branches? 😆

[formal mountain]

Yep! And thank you!

[quasi cobalt]

Happy Kitty

Sad Kitty

Saddest Kitty

little ball of fur

[uncut seal]

i see what you did there

and nice images too

[quasi cobalt]

lol thanks

[formal mountain]

this one looks so compelling. like subtle, but plaintive

[uncut seal]

purr purr purr ? 🎵

[quasi cobalt]

https://tenor.com/view/clapping-hands-yay-excited-applause-yes-gif-18387498984069029230

[formal mountain]

this image makes me truly happy

like a strawberry crossed with a cottonball

[lyric zephyr]

what r ur thoughts on the new Image Gen ?

[formal mountain]

I was skeptical of the concept until I realized that each preset has like new additional pipelines, just like we’ve had for alchemy and photoreal, but like a variety

Capybara! Capybara capybara capybara capybara, Capybara! (Emoji edition!)

[lyric zephyr]

yeah its like new Elements .... finally I got my Anime Screencap :KEKW:

I hate it that you cant type in Image Dimensions

[formal mountain]

yeah that might be something to put into suggestions

[formal mountain]

yes!!

[quasi cobalt]

I am worried that Dreamshaper V7 and other older models are going to dissapear

[formal mountain]

they will be available in the legacy mode

[lyric zephyr]

guess the training data is too old 😄

[quasi cobalt]

How long will legacy mode be available. Surely it's just temporary to soft-launch the new UI and let people get used to it.

[lyric zephyr]

i like Legacy more... some Combinations dont work with my current prompts

[formal mountain]

it will probably depend on user feedback. if people want to keep access to older models, they're going to try to accommodate that.

[quasi cobalt]

Oh OK lovely. I don't mind if newer models could replicate the same effects, but I do love a good dreamshaper v7 style image. 💖

Dreamshaper and Absolute Reality are my two faves. Obviously the XL models are absolutley amazing but sometimes there is a certain feel that you get with the older models. 🤗

[brazen lion]

Me too.

Will there be an option for customization of the aspect ratio?

[uncut seal]

this was one of the things documented during the demo earlier. hope so

I'm sure there will be an opportunity during my stream tomorrow to share my honest thoughts about the new UI. maybe it'll grow on me a bit by then 🤷‍♂️

[formal mountain]

Sorry, Thom. There's only going to be enough time to share any dishonest thoughts you may have

[uncut seal]

my opinion isnt for sale 😉

[formal mountain]

neither is our discord! hahahaha. but I guess I'll let you be honest this one time. then go back to being your normal, dishonest self

[lyric zephyr]

Character Sheet is pretty dope as pipeline

Image Gen v2

@flat ruin new Pipeline .... so I made examples 😛

[tiny galleon]

9.Cute Emotes ？

[tiny galleon]

Cute Emotes ？

[formal mountain]

hey! are you trying to create an image or asking about one that's been posted here?

[tiny galleon]

This combination of small picture

[pale lodge]

Hey, guys! I hope you find this useful

Made two versions for convenience 🙂

[formal mountain]

thank you Dimo!!

[lyric zephyr]

made this with new UI ... if anyone needs Wood 😄

[quasi cobalt]

https://discord.com/channels/989166677390426132/1243059835377946624

[tiny galleon]

I haven't used it yet

[quasi cobalt]

Well, have fun. It is very cute.

[quasi gust]

A reference image anyone can use. Left side reference and the right is the output result..

[lyric zephyr]

cant go there

[tiny galleon]

Ok

[lyric zephyr]

or use this 😄

[quasi gust]

Loll

[quasi cobalt]

it's not for you. It's only for Chinese speakers. 😉

[olive cliff]

hello Dimo
May I share these two pictures of yours with the Chinese community?🙏
I think it helps us a lot 🙇🙇

[pale lodge]

Hey)
Yeah, sure

[lyric zephyr]

still waiting to enter custom image dimensions again 😄

hmm whats your opinion about the "dynamic" preset from before ? since AnimeXL in the new UI has no more Dynamic ... which image share the same quality as with the generic Dynamic Preset

[formal mountain]

Legacy mode? 😁

[lyric zephyr]

in the new ui 😄

[formal mountain]

there's going to be some tradeoffs 😁 you might need to use legacy for some things, and creation mode for others

[dark sage]

This just sounds counter productive to me.
If they're not on par with each other...
They'll just have to maintain and upkeep both layouts individually.
Seems like a hassle and a headache. 🤷‍♂️

[formal mountain]

yeah, or they could bundle the legacy features in as advance mode options in the new interface

when we can make our own presets I think this will be less of an issue

they've heard feedback from everyone about the new interface, and they'll be adjusting things a bit, as well

hm, I'm curious

<@&1211672185069117451> <@&1220363318590705795> would love to get your thoughts. the team is genuinely trying to take in all the user feedback

and anyone else who would like to share their thoughts

[forest reef]

honestly there are few plus and minuses here and there but the most annoying thing for me is that negatives and models are placed inconviniently. While it is cool to have presets, the way its executed is confusing to me too...at first I thought it was model...after that I got confused by where alchemy was...another issue is that 765x512 cant be done hence its going to cost me way more than generating at legacy, right now for me atleast the minuses are bit higher to say completely that I like it, I do switch to enjoy presets and contrast however for everything else its pretty disruptive because fundamental placement of basics have changed drastically

[rose forge]

At first I was very lost, I'm getting used to it and finding where the tools are, but I'm very used to the previous system xD I think that's my biggest problem

I really liked that free users can also use the premium tools

(pd :Sorry if you don't understand if I'm in favor or I don't even understand xD)

[proven ridge]

When you use anime XL, for example, you shouldn't be able to use photoreal , so why is it possible on the new interface? And there are other inconsistencies

[shrewd adder]

My two Cents:
Overall I like the new UI. Once used to it it's easier to experiment around with all the styles, presets and settings.

Minor issues:

• I'd rather have it called Alchemy again instead of Quality, this way it's clearer that Alchemy represents a distinct technology instead of just "higher quality"
• Image Reference and Elements should be two separate buttons since they are totally different things

Major issues:

• the resolution settings are not very intuitive. I find it okay that you need to open a settings tab to adjust resolution, but I dislike the tall-wide-slider. I'd prefer the old sliders so I have direct influence on both input and output resolution
• the old models. I still like to use Dreamshaper for many things, and (especially for free users) it's way cheaper to use them while still having a high quality output

[uncut seal]

I agree with all these 👆 points. Especially the first, which I hadnt considered yet. Calling alchemy 'quality' now gives the impression that free users cant create good quality. Misleading.

[graceful badger]

i dont mind the slider, but id like to see the instant switch return..

can i use the text image input to change the text in this image? is that possible?

[graceful badger]

so you can use more than one image guidance? i didnt know that.. 🤔

What could be an idea, would be to put a text layer above the image, and the let ai "morph" the text into the image.

So like in this image. there's a lot of text in. if you could put a text layer above it, with the characters circa where the messed up letters in the image are, you would get a lot of control.

Would this be possible with the current guidance system?

[lyric zephyr]

New UI does not have MultiStyle sadly

[graceful badger]

using the text input didnt work at all. what i had to do to change the text, was to to bombard the prompt with NICOLEs.

[formal mountain]

yeah, that's one of the first things they'll be modifying

[uncut seal]

Has anyone used character reference to successfully get fantasy skin colours? red, blue, green..? tiefling, dark elf or orc style..?
likely trying that out tomorrow.. 🤔

[quasi cobalt]

Are you streaming tomorrow?

[tough mist]

for color ref, i prefer style reference. can get hair color, clothing, even eyes and mouth expression in some cases

[graceful badger]

to get bowstring correctly, use content reference with an image of an archer with a clearly visible bowstring 🏹

[rose forge]

F I think that arrow is going to rip off my nose xD when it comes out

Every time better

[uncut seal]

No, friday. But I could be persuaded to chill stream some of the prep. 🤔

[uncut seal]

I asked because a friend wants to see herself as a green skinned witch. Ive not tried it yet, but if its anything like fantasy skin tones from normal prompts, then it could be tricky.. 🤷‍♂️

Was hoping someone had experiences already

[rose forge]

More than the mistake of the bow and arrow, it is impossible to make a hammerhead shark

Submarino

[formal mountain]

One thing I’ve found is that, for trickier requests, sometimes explicitly stating the desired element from the style reference helps a lot. So like, you include the green and then also state that you want green skin.

[near haven]

I believe Ayumi has explained in the stream that the ratio slider is for optimization of generation so beginners won’t keep generating with XL or lightning models using smaller dimensions and complain about the poor results they get.

[near haven]

Text input should have worked. Did you use white text on black background for the image guidance? 🤔

[near haven]

AI surely has difficulty comprehending the anatomy of a hammerhead shark, let alone making a submarine out of it.

[formal mountain]

Sharks in general, they all look like some variant of Jaws unless you specify some descriptor like gentle lol.

I do believe this one was trying to be a hammerhead, and then at the last it lost its nerve and just added fins lol. Like when you keep getting a thing wrong and then you finally get it right but you mix yourself up and guess wrong again 🤣

[graceful badger]

yes, white text on black background, since the text on image was white. i can try to swap, black text on white background

[formal mountain]

hahahahah omg awesome

[quasi cobalt]

I think it's because the text is quite small and thin, whereas it usually responds more easily to heavy bold text.

[graceful badger]

it wasnt a problem adding the text. the problem was combining the input text with the content reference image. i tried a lot of different strengths with both the text and the image, but they just wouldnt combine..

[quasi cobalt]

ah ok. have a look at the ones on my profile I just did style ref on mid

[graceful badger]

ok, but this need to be done the specific image

[quasi cobalt]

I don't know what you mean.

[graceful badger]

you just copy the style, and adds text input to that, creating a new image, based on the original. but i need the text on the original image to change, and the rest to stay as is.

[quasi cobalt]

so you're not even using text to image?

[graceful badger]

i use image-to-image, and combine it with text input, but all i get are these kinda results :/

if i use content reference, i get this image

[quasi cobalt]

And what does the text on the original reference image say?

[graceful badger]

just nonsense...

[quasi cobalt]

I see.

so you managed to get what you needed by flooding the prompt with Nicole?

[graceful badger]

this was the original pic

[graceful badger]

yes

[quasi cobalt]

Good to know!

[graceful badger]

NICOLE a precisely detailed Alfonso Cuaron stylized blockbuster action movie poster, NICOLE in large fonts. NICOLE written with consistent letters. made with hard paper, computer graphics, and PVC. highly detailed surfaces. reflective, absorptive. NICOLE

with image-to-image

[quasi cobalt]

prompt is very powerful

[graceful badger]

then i got

i think 3 of 8 got the spelling right when using that technique

[quasi cobalt]

oh I see. well I guess that's pretty good odds

will you try text-to-image + content reference another time?

[graceful badger]

cant complain. still got 0 of nth on the correct bowstring withouth guidance o.O

oh, i done it already

right now im mixing up my prompt and style 🙂

[warm kestrel]

Hi Zarahuztra, I have been trying to make a submarine that looks like a bio-organic hammerhead shark for a longgg time now with no luck. Can I ak how you were able to do this? It is very close to what I have in mind, but I can't seem to even get Leo to take me this far so I am wondering what magic you were able to work lol thanks

[quasi cobalt]

He is in the Voice Channel at the moment, perhaps you want to jump in and join him there?

[graceful badger]

if you join the voice channel, ill can help you 😉

[warm kestrel]

Okay, thanks 🙂

[uncut seal]

in case anyone cares about my results with character reference prompting for fantasy race skin colours, I've gotten some great results in testing today.
seems to depend heavily on the colour though. Purple, Blue and Green, high success rate. Red, considerably lower rate of success. 🫤

[tough mist]

maybe try recolor in photoshop or something then try put then as reference again

[uncut seal]

too complicated. i dont wanna teach people 'too many steps'. i wanna keep it as simple as possible. if red dont work, then I'll just go with blue, green and purple 😁

[quasi cobalt]

Best to show everything that Leonardo is capable of for the Leonardo community rather than recommend a third-party application.

[uncut seal]

that too

[tough mist]

I’m being pragmatic here and my point is to solve the problem and use Leo tool if it is available within their suite 🙂

[uncut seal]

and there's nothing wrong with that 👍

[formal mountain]

Does red just fail to show up or is it very faint?

[uncut seal]

Some of both. More of the first and less of the latter

[formal mountain]

Do you specifically say "red skin" in addition to the prompt? and you're using style refrence or char or both? I've had some success with the faint result back in as a content reference image in addition to the style ref image

[uncut seal]

All Im doing is character reference with a normal foto. No style, content, etc.

So yea, red skin or red skinned, with and without hyphon, tried a lot.

But no biggie. Feel free to try too

[formal mountain]

I would suggest doing style reference with the red you're aiming for

[uncut seal]

Trying to keep it simple. Wanna get the max out of just prompting before i resort to multi-reference or external tools

[formal mountain]

there's a ton you can do before external tools, but multi reference really enriches things

[lyric zephyr]

Any Infos about how the image scale on new UI (small/medium/large) affects details ?

[graceful badger]

whats the deal with adding text to image? i get really bad results, both using content ref or image-to-image. this is the best image ive gotten so far..

this is the original

From this

to this..

[quasi cobalt]

https://cdn.leonardo.ai/users/fa466aaa-cdd2-46ea-a529-92bc0a96898d/generations/bf555d64-bd54-4710-8782-b6800adb09d5/Default_An_art_exhibition_walls_lined_with_bold_coloured_paint_1.jpg

[graceful badger]

how do you add text to images?

[tough mist]

multiple references are conflicting by nature

[near haven]

If you are trying to use content reference together with text input/ edge to img/ depth to img to generate the word over the img, there will be a problem as the AI will prioritize the content reference over the text.

[quasi cobalt]

Let the record state that I did not ask for the image to say those words.

[quasi cobalt]

I use text-to-image.

[graceful badger]

i use either text input + content, or text input + image-to-image. none work very well :/

[quasi cobalt]

which model?

[graceful badger]

its better to use the prompt to change the text, even add text to the prompt to enhance the correct spelling

kino, lightning, anime, or whatever?

[quasi cobalt]

Albedo and Kino are usually best for text ( I don't know anout with new UI )

https://cdn.leonardo.ai/users/fa466aaa-cdd2-46ea-a529-92bc0a96898d/generations/286601a8-5db9-4ba6-903d-595b49ff49e6/Default_White_tEXT_NICOLE_Graphite_illustration_of_a_beautiful_0.jpg

[graceful badger]

whats your ratio of correct spelling?

[quasi cobalt]

100%

you have prepare a good text reference image though.

[graceful badger]

style reference?

[quasi cobalt]

The other day you were saying the text was part of your image, rather than being seperate text-only file

[quasi cobalt]

Text Image Input

See my screenshot, you can see I used two input images. one for the text, then another for style

[near haven]

I

[quasi cobalt]

Once I wrote in the prompt something different to what my text file said and it just combined the two

[graceful badger]

ok

[graceful badger]

but you, like me, lose colors

[near haven]

That was done with Content reference High and Text Input at 2.0

[near haven]

That's kind of unavoidable as the two image guidances are conflicting with each other.

The easiest way I'd go for is to use Canvas

[quasi cobalt]

Mine are white text on black background

[near haven]

If I want to retain everything in the image

[quasi cobalt]

@graceful badger look at my examples in #1163531400021024889

[near haven]

or use other software to edit in the text

[graceful badger]

but what if there's already text there? and you just want to change that text?

then you bombard the prompt..

[near haven]

At 1.5, the text can be there but by chance.

[near haven]

You can use Canvas to edit that quite easily

[quasi cobalt]

Then you don't have enough clarity for the ai to see what's going on. You can still do it, but the success rate would be lower

Yes, I agree, in that case you can just use canvas editor to change the text, and keep everything else in the image the same

[graceful badger]

why was photoread with anime removed?

[near haven]

All the texts on the right are edited in with Canvas

[graceful badger]

image-to-image + adding a short word multiple times in the prompt, alongside words like font, and lettters, creates a semi-high ratio of correct words

you should be able to do canvas inside the image generator

i do use canvas if i get images i really like, but needs a bit of fixing 😉

[near haven]

Fun fact (if you haven’t learnt this already): In text input, the AI treats white colour as the foreground and black as the background.

What you prompt for normally with text input is the style of the text you want, and what you prompt for the background will show in the black area.

In order to keep the image you prompt for like Zara’s case, it’s better to use white background and black text so that what you prompt for will go to the white part to be shown.

[near haven]

If I use black background and white text for these examples, the number will not show at all.

[quasi cobalt]

That is interesting.
In one case however, I asked for the text to be engraved, and it looked like it was indented. But in other case when I didn't say anything, then it looked three dimensional.

[quasi cobalt]

Similar to my concerns about the old models being lost, another perosn in General is hoping 3D Animation Style wll not be lost as well
I think it's fine if it can be closely replicated, but that's a really specific style that I expect many people will miss.

[formal mountain]

The legacy mode is intended to stay, to provide users with access to the classic Leo tools. there may be new features which only show up in the new interface, but there are no plans to remove the ability to keep using older models.

we've already seen that some features aren't compatible with older models, and for this reason they aren't being included in the new interface, in an effort to simplify use for less experienced users. but the policy remains that we will keep the older models for expert users to access.

[formal mountain]

The Pretzel Stand

[lyric zephyr]

hey its me xD

I made you a random Gemstone Texture xD

turned into this 😄

[formal mountain]

fun!

[lyric zephyr]

just made random circles in Photoshop and used them as content reference .D

[formal mountain]

sorry I heard "I made random circles in canvas" 🤣

[lyric zephyr]

Imagination is unlimted 😄

[formal mountain]

does it work with this for content ref?

[lyric zephyr]

I am pretty sure haha

[formal mountain]

hahaha so that was "random circles" in realtime gen 😂 now I want to see it in gem form 😂

[lyric zephyr]

for start take this

[formal mountain]

ooh pretty

[lyric zephyr]

use if you want 😄

[formal mountain]

nice!

yay!

[lyric zephyr]

#1235650818678456380

[formal mountain]

should make a game of this, assembly line haha

one person makes the content ref image, the next person turns it into style ref image, next person makes final product lol

[lyric zephyr]

well what if the content ref. image is .... one pixel ?

people can easy troll :KEKW:

[formal mountain]

then your mileage may vary 😂

[formal mountain]

people can always troll lol

[lyric zephyr]

because its fun hehe

[formal mountain]

wait, Danny, are you admitting to being a troll? 🤣

[lyric zephyr]

I am in the Troll-Spectrum 😛

[formal mountain]

awesome 😂

care to use your powers for good, and attempt the #1247105139341393951 challenge?

[lyric zephyr]

have to prepare for Denis LUL

[formal mountain]

oh you have your interview? well why are you here trolling? 😂 is trolling how you prepare? 🤣

[lyric zephyr]

may... be .....

btw. if you use a different animal and prompt for a different animal

u get funny stuff

[formal mountain]

hahaha omg awesome!

[lyric zephyr]

make bunny content ref and prompt for a different animal :KEKW:

[formal mountain]

you'll get a pooka

[lyric zephyr]

seems with sketches u get kinda toys xD

here is a @formal mountain based on alpaca picture :KEKW:

hahaha

[lyric zephyr]

@formal mountain here is something for an expieriment 😉

[formal mountain]

What model did you use for the second image? Kind of reminds me of an image I made once lol

[lyric zephyr]

Preset chose Albedo 😄

its nice how such lowres images can affect the outcome 😄

[formal mountain]

#1082196308699398165 message

My first ever image shared on Leo discord

[lyric zephyr]

with simple methods hehe

[formal mountain]

It was the Superhero Challenge, and I made flaming dragon lady 😂

[formal mountain]

I think this is her in casual clothes 😂

[lyric zephyr]

😄

new texture experiment 😄

seems like an ethereal blend of colors

kinda reminds me of Glowwave

[formal mountain]

used this one as style reference, to make this:

[lyric zephyr]

pretty

[formal mountain]

I'm getting some really sick results from this

@lyric zephyr use this one:

[lyric zephyr]

awww birbs

[formal mountain]

this is one of your first images shared here!

[lyric zephyr]

here is kinda like ..... Space-Themed lowres Image xD

MID

Ultra .... hmm

[lyric zephyr]

here is an interesting texture 😄

[formal mountain]

@lyric zephyr did you try using this as a style ref image?

[lyric zephyr]

just made this

here is your bunny 😄

here are some bubbles 😄

[formal mountain]

are you posting these in your textures thread?

[lyric zephyr]

na

[formal mountain]

couldja? 😂

it's more useful to people long-term if it's there rather than here

[near haven]

Has any one of you experimented with negative value of the Glitch art Element? 🤔

[formal mountain]

oh not yet! I had found mixed results with some of the recent elements in negative, I'll check it out now!

[near haven]

I have had a hard time generating exaggerated giant weapons on Leo.
Using a negative value in Glitch art seems to shut down the “size limit”, suddenly making the prompt work.

[formal mountain]

coooool

Incoming Cloud Strife! lol

[lyric zephyr]

OMG

same ... but -1 Glitch xD

[near haven]

Is this big enough for you?

[lyric zephyr]

nice one

try giant fists

[formal mountain]

weapon and umbrella!

[near haven]

Hmm… maybe this is a bit too much…

But size matters!

[formal mountain]

the cat's just like "yeah I'm out" hahaha

[lyric zephyr]

how about two weapons ?

[near haven]

Make it three

[near haven]

Are these big enough for you?

[formal mountain]

oh wow sweet

[near haven]

They need quite specific setting to show though… It’s pretty hard to override the presets which tend to generate perfect proportions.

[formal mountain]

yeah, sometimes I feel like my choices in elements and prompts are about fighting the AI haha

make it do what it doesn't want to do lol

[near haven]

I think while leo’s presets aim at perfecting certain qualities, it unavoidably hinders creativity and variation at the same time…

[formal mountain]

very definitely.

I like finding keywords that sort of allow it to break out of that

for instance "double exposure"

I'm not completely certain Leo fully understands what double exposure means, other than that it breaks rules lol

[near haven]

Oh, about “double exposure”, it definitely works less well on Lightning models than in the other XL models. I think it does best on Kino.

Can’t ask for triple exposure though

[formal mountain]

ah yeah I tend to like to use Kino. Probably many of us have that model that we keep going back to, lol

[near haven]

I am wondering if there will be some sort of “soft lens/focus” element that gives dreamy atmosphere with high quality. Right now it’s pretty hard to pull that off with the current alchemy/ quality setting.

[formal mountain]

ahhh yeah that would be nice

[near haven]

While “White Ethereal” was out, I thought that was it, but in fact not quite.

[formal mountain]

does it get part of the way there? or do you have any tricks?

one of the challenges for the team that works on Elements can be that if it's hard to create the image, it can be hard for them to build a dataset.

[near haven]

Not quite

Lambent, High key lighting, soft-focus filter, dreamy glow, photorealistic digital illustration of a cute anime Easter Goddess warmly smiling, white modernized costume, Key Visual,maxmalism, light particles, unique artistic composition, whimsical magic, elaborate consistent background elements, Key Visual, cel shading, Rembrandt lighting, bokeh, Siloa

This is without alchemy or elements. You can see I've put in as many descriptors for the "soft lens" effects I want in the front.

This is with no alchemy with WE 0.3 + SPA0.3

I added "sfumato" in the prompt and "outline, solid lines" in the negative.
And dailed WE & SPA both up to 0.5

Once I put alchemy/quality in, it gets sharper.

[tough mist]

amazing, just by prompt?

[quasi cobalt]

Is thisone free for anyone to use? the colours are gorgeous! 😍

[lyric zephyr]

like any texture I post 😄

[quasi cobalt]

awesome!
Pictures coming soon .... 🤗

[formal mountain]

wow!! this is so lovely

[quasi cobalt]

the colours with the gold trim worked really well

The contast between the subject and the background is perfect. 😍

[formal mountain]

omg yay mini robot!

kind of my favourite thing to gen lol

[formal mountain]

I Wish We Had An Emoji For…

[near haven]

Yes, of course

[digital walrus]

@pliant crest 👀

[pliant crest]

like bokeh?

[near haven]

Bokeh is for the background only though.
It’s more like a high key lighting/ soft focus/ soft lens effect

Kind of like this

[near haven]

Got some more reasonable giant fists

[tough mist]

looking forward for you guideline

[near haven]

It’s doable but not stable though.

[uncut seal]

What for? 😲
Prompting for Vi from league of legends..?

[near haven]

It’s a request from Danny

[lyric zephyr]

Have you tried giant pistol ?

[uncut seal]

What foor? 🤨

[lyric zephyr]

Not easy to make it not look like cannon

[near haven]

It's bigger... let's see if it can get even bigger

[uncut seal]

big-ish pistols

[lyric zephyr]

Or the opposite..make someone hold a medieval cannon or ballista in the hand

[lyric zephyr]

I guess creating low quality reference and use them on mid is my theme now xD

[formal mountain]

a trebuchet 😂

plink.

[wide pebble]

@quasi cobalt helped me go from a basic image/idea that didn't reflect my colors and the style of my environment at all, to the creation and reproduction of a wallpaper as I wished. He accompanied me step by step, was very pedagogical and saved me precious hours of trouble trying to tinker on my own for a result that would surely not have been as good.

Long live Willow !

❤️

😂

[formal mountain]

yay willow!!!

so happy to hear this!!

[forest reef]

3koma gag comics

[odd forge]

Hi friends\

[formal mountain]

Okay so you are wanting to take this mockup image and make it pop?

[odd forge]

YUP

I have many other designs but want to have something really viable so developer can use the images to build the theme

[formal mountain]

Okay so you’re working with an app developer and want to communicate with them what you envision?

[formal mountain]

how familiar are you with canvas editor?

[lyric zephyr]

Low Resolution Texture

[wind yarrow]

Hello, can you tell me where is the qr code to image?

[formal mountain]

Hi there! To use the QR code feature, please select a non-XL model and disable Alchemy 😁

[ripe wind]

Hi everyone, I am battling to get the settings/prompt correct to get scary, shining eyes from the forest at night looking at the character, any ideas perhaps?

[lyric zephyr]

looking a viewer, glaring eyes, dark background, high contrast ... anything can help 🙂

[formal mountain]

So like the look of a dark forest with little dots of scary eyes from the bushes?

[formal mountain]

I see what you mean, that's tricky. I got close but it doesn't want to put the eyes in pairs

we'll find a way!

meanwhile, this is a hilarious fail:

[formal mountain]

I kind of got the look in realtime canvas

[ripe wind]

@formal mountainIt does seem scary though so it fits a part of the prompt.....\

I'm also trying in real-time canvas, but it sorta defeats the object of having an AI system that generates images but you still need to do the nitty gritty stuff.

[formal mountain]

yes definitely. my thought was to perhaps get a good style reference image

[ripe wind]

@lyric zephyrRun Forest, run gets a whole new meaning with that creation of yours!

[formal mountain]

the biggest challenge seems to be getting it to do the eyes in pairs

[formal mountain]

using this as a style ref image helps get them into pairs, but results in some hilarious misses:

it really wants to give form to the creature behind the eyes

bush-creatures, trying to blend in with the foliage

not so hidden:

@dark sage eeriness and eyes, that's kind of your wheelhouse. ever attacked this challenge before?

[uncut seal]

did you specify a colour or palette you wanted?
not gonna go thru and read it all again, but everyone seems to have gone with red..?

[ripe wind]

@uncut sealred and yellow, even shiningwhite will work

[fallow spoke]

Hi there. Does anyone knows how we can train a model to get consistency on character at this new version?

[formal mountain]

hey! there is the option to use character reference, or if you switch to legacy mode you can access your custom trained models. many of us are finding, though, that character reference is enough that we don't need to use the custom models

[fallow spoke]

Thanks #pookah

[formal mountain]

happy to help! let us know how it goes!

[fallow spoke]

I will!

[dark sage]

Are you looking for an images of disembodied eyes specifically, in pitch blackness?
Or do you want a group of creatures staring at the camera with glowing eyes?

https://tenor.com/view/eyes-black-black-and-white-look-around-looking-around-gif-10347371540908890270

[rose forge]

XD

[ripe wind]

@dark sage I want the eyes only like when you shine a flashlight on it.

[rose forge]

Style Jon Klassen and Rafael Albuquerque art.Eyes of a group of cats in the dark.photo by Bernie D’Andrea, glowing retinas, group of eyes growing on creatures, with glow on some of its parts, eyes are yellow.

[quaint blaze]

Heyho, i would love to create ukiyo-e paintings, what settings should i use?

[uncut seal]

Well first Id say the prompt is more important than the settings. Have you tried to create any yourself already?

Someone was playing around with that style not tooo long ago. Lemme see if I can find it..

[quaint blaze]

yeah i tried but if i describe the ukyoe painting style it doesnt really take over the promts and instead for example draws the landscape i described it as

[uncut seal]

paintings of what, are you hoping to make btw. what would be the subject or landscape you're aiming for?

here, i just searched the key word ukiyo in the searchbar. looks like it was @flat ruin who was last playing around with it. #🙋│image-help message

maybe he's made some more experiences since then, or maybe you can grab something from the conversation at that time, that's useful..

[quaint blaze]

thanks for the help y try to figure it out 🙂

[forest dragon]

Ooooo what is this place

[uncut seal]

heck, he found us..
its where the yellow-named people hang out a lot. check the sub-threads 😉

[ruby wraith]

https://tenor.com/view/imagination-land-south-park-imagination-gif-5897929

[forest dragon]

Ahhhh okay interesting

Like a mentor program?

[uncut seal]

no thats another other channel. this is kind of a more fun image-help. if you can imagine such a thing

[formal mountain]

#📢│community-updates message

😁

[formal mountain]

this will likely make the task much easier: #📢│announcements message

[lyric zephyr]

I really like the new model 🙂

[formal mountain]

yeah it's addicting!

[rose forge]

Yes

I liked too

[formal mountain]

they look hungry 🤣

https://tenor.com/view/little-green-men-aliens-three-eyed-squeaky-toy-aliens-toy-story-a-stranger-from-the-outside-gif-17234188

[rose forge]

The three eyes were very complicated

[formal mountain]

the claaaaaaaaw 🤣

yeah crazy it got the eyes!

[formal mountain]

it's breaking out of "logic" and assumptions, the limiting biases

[formal mountain]

awww, I love you too, tiny aliens 🤣

[formal mountain]

this is lovely! have you been using the new phoenix preset?

[quaint blaze]

nope i just saw that it is available ,will try

[ripe wind]

@formal mountainThx it works much better!

[formal mountain]

yay I'm so happy to hear this!

[ripe wind]

Is it possible to add a graphic designer on my acc in Leonardo? Can I just use my log-in details and can both of us use it simultaniously?

[formal mountain]

we'll be adding team plans in the near future. for now, register your interest in paticipating in a trial of teams, here: https://leonardo.ai/team-plans/

[quaint blaze]

oh damn

[ripe wind]

@formal mountainok thx I did do it now, in the meantime is there a way around it?

[quaint blaze]

That is actually crazy, faces are a bit bad but damn is that intersting to see

[maiden tide]

✋

[jagged moat]

I am trying to create more assets for my game matching the same style and color of our current assets. Is there someone that can help me? i cant get it to match the correct colors. Ive tried upwork ive had multiple people try and cant find someone

[near haven]

Have you tried using style reference?

[jagged moat]

yes for weeks ive been trying and i trained a model off the current assets. and ive tried using seeds and the style reference needs to be the exact color?

im trying to do assets like this i dont even know which model to use

[formal mountain]

Okay so you’re wanting this art style but different colors and shapes? Did you generate the originals with Leo?

[jagged moat]

no we did the originals with blender. but i have one icon in leanardo i did and i try to use the seed and it doesnt work

[formal mountain]

Okay. What are you looking for as the end result? Like what type of objects and variations? Are you looking for foreground elements, power ups, platforms, interface, other?

[jagged moat]

more assets which look like they belong to the set stuff like this in the same style and texture. if i buy a large package is there someone that can help me set it up?

[forest dagger]

Does something like this works for you?? (style)

[jagged moat]

no it needs to be the same color and texture and style as the first image this is not as plastic looking and too detailed and the wrong colors

[rose forge]

Mmm

And if you want the same color, the same texture as the first image, why don't you use that same image?

[jagged moat]

the riddim image it was made in blender it is our image we are using. but we want other assets that match. we want to make a crate and a person in the hood the syle should be consistent the crate is not the same style as ours

[rose forge]

I'll see what I can create :3

[jagged moat]

that would help me allot. are you available to hire if you can do it?

[uncut seal]

Do you have a paid membership or are you on the free plan of leonardo.ai?

[jagged moat]

paid

[uncut seal]

It sounds like just using one or multiple of your existing assets in style reference should do what you need..

Since style copies your colour palette and textures.

As for model and presets.. Id try Leo Diffusion XL to start. 🤷‍♂️

[rose forge]

It's like he says use image character+ image content

[jagged moat]

this is what i get

this

[rose forge]

Image character (My photo)+image content (image 1)

Final result image 2

I think the result is better with that combination.

To get something similar

But adding content

[near haven]

The reason why you get those results is that the AI was trying to create a root box that looked like the box for content reference but with smooth edges and color scheme like the other two images you used for style reference.

You probably have to be a bit more specific in what you have in mind to help the AI create the correct image.

For instance, without specifying what is in the lootbox, the AI may not generate "coins" in the box properly.

[tough mist]

1. try decrease style weight. I think the model are adhere to color too much from your liking I suppose
2. add 0.2-0.4 weight of simple icon or flat illustration element seems compatible with tour style
3. try other model. I generally recommend AnimeXL for its cost and variety

[shrewd adder]

Coins, anyone? 😄

[formal mountain]

wow!!

[shrewd adder]

I'm using them for roleplaying now! 😄

[formal mountain]

omg awesome!

[shrewd adder]

Yeah I love it! It's so cool to use AI created stuff in my favorite hobby! 🙂

[forest dragon]

hey malakai

long time no see

the meme variations be endless now

[shrewd adder]

Oh yes, they are 😄

[forest dragon]

all time F1 memes

[shrewd adder]

I'm not into F1 ^^
context?

[forest dragon]

Just a meme. A driver crashed and he called himself stupid

[formal mountain]

Ahhh he said it about himself lol. I was confused, he seems like a nice enough guy, why are we calling him stupid 🤣 😂 now it makes sense 😂

[forest dragon]

https://youtu.be/aS4Me48wayM?si=hlPrcPbO7GYdUy-L

I guess I need to explain haha

[shrewd adder]

ouh... not good 😄

Hey guys, anyone here already tried on HDRI 3D projections?

[forest dragon]

on Leo?

[shrewd adder]

yes

[forest dragon]

i haven't... but i'll try now

i'm assuming HDR preset + 3D scult element?